Non-maximum suppression (NMS) is an indispensable post-processing step in object detection. With the continuous optimization of network models, NMS has become the ``last mile'' to enhance the efficiency of object detection. This paper systematically analyzes NMS from a graph theory perspective for the first time, revealing its intrinsic structure. Consequently, we propose two optimization methods, namely QSI-NMS and BOE-NMS. The former is a fast recursive divide-and-conquer algorithm with negligible mAP loss, and its extended version (eQSI-NMS) achieves optimal complexity of O(nlogn). The latter, concentrating on the locality of NMS, achieves an optimization at a constant level without an mAP loss penalty. Moreover, to facilitate rapid evaluation of NMS methods for researchers, we introduce NMS-Bench, the first benchmark designed to comprehensively assess various NMS methods. Taking the YOLOv8-N model on MS COCO 2017 as the benchmark setup, our method QSI-NMS provides 6.2× speed of original NMS on the benchmark, with a 0.1% decrease in mAP. The optimal eQSI-NMS, with only a 0.3% mAP decrease, achieves 10.7× speed. Meanwhile, BOE-NMS exhibits 5.1× speed with no compromise in mAP.
@article{arxiv.2409.20520,
title = {Accelerating Non-Maximum Suppression: A Graph Theory Perspective},
author = {King-Siong Si and Lu Sun and Weizhan Zhang and Tieliang Gong and Jiahao Wang and Jiang Liu and Hao Sun},
journal= {arXiv preprint arXiv:2409.20520},
year = {2024}
}